Context-Sensitive Augmented Reality

1. A method comprising the steps of: obtaining an output of an image sensor, the output being representative of a real-world environment comprising at least one real-world object; determining, based on the at least one real-world object within the output, information representative of at least one weather condition; inserting a virtual object into the output representative of the real-world environment; analyzing the information representative of the at least one weather condition using a neural network to determine meta data representative of at least one modification to the virtual object; modifying the virtual object using the meta data, such that the virtual object will appear to be affected by the at least one weather condition; preparing the at least one modified virtual object for display; and displaying the modified virtual object on a display.

2. The method according to claim 1 , wherein the information representative of at least one weather condition comprises any of: wind direction and magnitude; and whether it is raining, hailing, or snowing.

3. The method according to claim 1 , wherein the information representative of at least one weather condition is based on a predefined region of the output of the image sensor.

4. The method according to claim 3 , comprising obtaining information representative of at least one weather condition for a plurality of predefined regions to determine a plurality of modifications to apply to the at least one virtual object.

5. The method according to claim 1 , wherein the information representative of at least one weather condition further comprises depth information obtained from a depth sensor, mono-depth convolutional neural network, or microphone.

6. The method according to claim 5 , wherein the step of analyzing further comprises using the depth information to determine whether the virtual object is partially obscured.

7. A device comprising: an image sensor for capturing a representation of a real-world environment comprising at least one real-world object; a display interface for outputting the representation to a display; and at least one processor for modifying a virtual object; the processor arranged to: determine, based on the at least one real-world object within the output, information representative of at least one weather condition; insert the virtual object into the representation of the real-world environment; analyze the information representative of the at least one weather condition using a neural network to determine meta data representative of at least one modification to the virtual object; modify the virtual object using the meta data, such that the virtual object will appear to be affected by the at least one weather condition; and provide the modified virtual object to a display controller for outputting to the display.

8. The device of claim 7 , wherein the at least one processor is any of an image signal processor, a neural processing unit, a central processing unit, and a digital signal processor.

9. The device of claim 7 , further comprising a depth sensor for obtaining depth data associated with the real-world environment.

10. The device of claim 7 , wherein the information representative of the at least one weather condition comprises any of: wind direction and magnitude; and whether it is raining, hailing or snowing.

11. The device of claim 7 , wherein the processor is further arranged to obtain depth data from the representation of the real-world environment.

12. The device of claim 7 , further comprising storage for storing the information representative of at least one weather condition.

13. The device of claim 7 , wherein the processor determines the information representative of at least one weather condition based on a predefined region of the representation.

14. The device of claim 7 , wherein the display is a head-mounted display.

15. The device of claim 7 , wherein the display is a transparent display.

16. A non-transitory computer readable storage medium comprising a set of computer-readable instructions stored thereon which, when executed by at least one processor, cause the at least one processor to: obtain an output of an image sensor, the output being representative of a real-world environment comprising at least one real-world object; determine, based on the at least one real-world object within the output, information representative of at least one weather condition; inserting a virtual object into the output representative of the real-world environment; analyze the information representative of the at least one weather condition using a neural network to determine meta data representative of at least one modification to the virtual object modify the virtual object using the meta data, such that the virtual object will appear to be affected by the at least one weather condition; preparing the at least one modified virtual object for display; and display the modified virtual object on a display.